Powerpoint compatibility: text not displayed identically even when substitute fonts have same metrics

This pptx [originally MSO generated, having Calibri font] has text displayed differently when opened in LO even when substitute fonts [Carlito, since I don’t have Calibri] have same metrics. Due to same metrics, it should have been exactly similar, I expect.

I’m uploading pics of all slides, how it appears on MSO online version. One can open doc in LO and compare with the pics. See that at the end of one/multiple lines, one/few words either have been pushed to next line or extra words have been included in the line: 11 20 24 25 27 30 35

I’m uploading this side by side comparison for only few of the slides via gif


Build ID: 8061b3e9204bef6b321a21033174034a5e2ea88e
CPU threads: 4; OS: Linux 5.4; UI render: default; VCL: gtk3
Locale: en-IN (en_IN); UI: en-US
Calc: threaded

Out of curiosity I had a quick look side by side LO and MS Powerpoint 2010 on same PC. I think most of the differences are caused by how each program fits the text to the space. Use the free Powerpoint Viewer for pptx I guess.

If I untick Autofit text in LO that starts to reduce some of the difference. In Powerpoint the line spacing is often reduced to 80% or 90% in these slides, this might be a programmed method of sizing text or might be a manual override by the creator of the slides. Changing LO line spacing to match and they both look more similar, as does exactly matching font sizes. There are still differences though, in Slide 5, LO substitutes “ti” and “fl” with ligatures in the heading whereas PPT 2010 does not. The underlining in LO is further below the base of the text then in PPT, subjectively looks better in LO. The Calibri font renders with slightly softer edges in PPT compared to Impress; if I change them both to Carlito that difference disappears

reported at fdo#136231

The expectation of compatibility is generally very high.
Unfortunately this cannot be fulfilled in all cases.
For one thing, it can be a problem with the different fonts (same is not always the same).
On the other hand there is no 100% compatibility between M$ and LibreOffice.


Always create and save your files in LibreOffice and save them in ODF format (ODT, ODS, etc.).
Always keep these files as their source. If you need other formats for distribution to partners, you can open an ODF file and save and distribute another format with ″ Save as… ″.

This way, you always have working files available in your system environment.



Please report the behavior as a bug in Bugzilla .

Please announce the link from the bug here.

To do this, edit your initial question. Thank you.


Due to same metrics, it should have been exactly similar, I expect.

“Exactly similar” is an oxymoron. Your expressed expectation is exactly what you got, which is different from what you actually expected.


Best practice already suggested by @Hrbrgr. Rebuild your presentation to work in Impress. Save as LibreOffice presentation.

Posed question

What are the exact factors playing here which is causing even same metrics font to have different display?

The short of it

You will most likely never find a perfect answer for that question. (See my elaboration - “The long of it” - below.)

The fundamental cause, which can be used as a working rule of thumb, is that Microsoft Office and LibreOffice adhere to different “schools of typography”, and in particular they handle spacing differently.

  • Some object anchoring settings have different outcomes in some cases.
  • Edge detection (where character glyph meets box edge or text margin) seems to be slightly different.
  • MS Office works harder to “read your mind” and adapt content to available space.
  • Etc.

The long of it

You have introduced more than one variable, and additional variables may be introduced from settings in the applications. Most likely it is a combination of factors.

  • Variable 1: Different applications
  • Variable 2: different fonts
  • Variable 3: different system (other operating system running on other hardware).

Settings for kerning, the use of default styles and the differences in styles handling in general, automated adaptation and possibly other factors which elude me at the moment, complicates the picture further. In any case, determining the exact factors requires some work.

  • For the differences between fonts, Earnest Al has done a bit of the work for you. To determine the exact impact from font differences, you need to compare the use of the two fonts on the same system running the same software. I gather that a research environment for this is unavailable to you at this time. I have that, but I don’t have the hours to spend for such a fruitless task.
  • To determine the impact from using different systems (where one or the other app may adapt the rendering according to screen geometry or resolution) you need to run the same presentation in the same app on both systems.
  • To determine the impact from differences in programming, you need to run the same presentation on the same system with different apps.

Another way to determine factors would be to inspect the program code and font data, and do a “stepped run” (by hand or by variable inspection at carefully set breakpoints) to determine how the rendering develops. This is an exact method (whereas the above methods use “circumstancial evidence” to support a best guess) but requires significantly more resources (time, skills, and access to programming sources for applications as well as fonts.) This will only be practically possible for members of the Microsoft Office development team, because they don’t subscribe to the open source philosophy so their coding is only available to insiders.

Thanks, now i know possible reasons for it.